Panel-based digital audience measurement — Pakistan
YouTube reports what happened on the channels you own. Padiam measures what a consenting sample of Pakistani viewers watched across the whole platform — and which advertising actually reached them.
Most viewing lands outside the catalogue
A number you cannot defend is worth less than no number at all. So Padiam measures what people did, not what they remember doing. It counts Shorts and long-form as separate units, because YouTube does. And it withholds a reading the sample cannot carry, rather than printing it with an asterisk.
Consenting panellists install a metering app and then forget it exists. No diaries, no prompts, no recall. It records what played, for how long, and what advertising appeared beside it — with age, gender, region and socio-economic class already attached from enrolment.
Recall tells you what a person remembers. Metering tells you what happened.
In parallel Padiam counts the platform itself — public figures for every catalogued channel, collected daily and versioned so that history is never overwritten.
Census gives you totals. The panel gives you people. Padiam reconciles them and reports them separately, because a complete count and a sample estimate are different kinds of fact, and averaging the two produces a number that means nothing.
On the device, Padiam observes the advertising a real viewer was actually served: the format, the advertiser, how long it held the screen, whether it was skipped, and where it fell in the break.
A delivery report tells you what was served. Padiam tells you who was exposed, to what, and for how long — measured at the person, which is what a currency has to be.
Nothing renders without provenance. Each number declares whether it is a census count, a panel estimate or a derived value — and if derived, the formula that produced it, carried with the figure rather than filed in an appendix.
Readings the sample is too thin to support are withheld, not flagged. A footnote survives being read and disappears the moment a column is sorted.
A fixed panel answers fixed questions. Padiam can compose a panel around a defined one — a category, a city, a language, a campaign window — meter it for exactly as long as it matters, and stand it down.
Continuous measurement where you need a currency. A focused sample where you need an answer.
Q.01 — Audience
Platform analytics describe the people who arrived. A panel describes the population they were drawn from — including the ones who went somewhere else.
Q.02 — Exposure
Measured on the device, at the moment of exposure — not modelled backwards from an impression count.
Q.03 — Share
Because the panel is not tied to any one channel's account, it can compare surfaces nobody in the room owns.
Q.04 — Discovery
The most useful thing a panel finds is usually the thing that was never on the list. A census can only ever count what it was pointed at.
Platform data
Views, watch time and impressions for the inventory you bought, on the surfaces you own. Complete and exact — and unable to tell you who those viewers are, what else they watched, or what a competitor's advertising reached them with.
Padiam
Which people watched, what else they watched, who they are, and which advertising actually reached them. An estimate, and labelled as one — projectable, comparable across channels nobody owns, and defensible line by line.
We would rather walk you through how the numbers are made before showing you the screens they appear on. Ask, and we will.
Email padiam@cincorios.co